三峡梯级生态友好型多目标发电优化调度模型及其求解算法  

TGC eco-friendly generation multi-objective optimal dispatch model and its solution algorithm

作　　者：王学敏[1] 周建中[1,2] 欧阳硕[1] 张勇传[1,2] 

机构地区：[1]华中科技大学水电与数字化工程学院,湖北武汉430074 [2]数字流域科学与技术湖北省重点实验室,湖北武汉430074

出　　处：《水利学报》2013年第2期154-163,共10页Journal of Hydraulic Engineering

基　　金：国家自然科学基金重点项目(51239004);高等学校博士学科点专项科研基金(20100142110012);水利部公益性行业科研专项(201001080)

摘　　要：梯级枢纽生态调度是当前国内外河流生态恢复与重建研究中的热点问题。本文基于逐月频率计算法并结合长江流域相关生态要素,制定了宜昌站适宜生态流量,作为生态效益评价标准;然后针对长江中下游主要生态问题和三峡梯级枢纽具体工程实际,结合生态效益评价标准构建了梯级枢纽生态友好型多目标发电优化调度模型。为有效求解多目标优化问题,提出一种包含外部种群的双种群多目标差分进化算法,并设计了精英选择和混沌迁移机制实现两个种群间的信息交换,提高了算法的多目标优化性能。三峡梯级枢纽实例应用研究表明,本文算法能在较短计算时间内获得多个符合生态效益评价标准、分布均匀、收敛性较好的非劣调度方案,为制定合理的调度方案提供了科学的决策依据。The optimal ecological flow of Yiehang station is determined as ecological benefit evaluation stan- dard based on monthly frequency computation method and ecology elements of the Yangtze River. An eeo-friendly generation multi-objective optimal dispatch model of cascade reservoir is proposed according to the major ecology problems and the specific engineering practice in the Three Gorges Cascade （TGC）. A dual-subpopulation multi-objective differential evolution algorithm with external archive, elitist selection and chaotic migration is presented to solve the optimal dispatch model. The results reveales that a lot of non-dominated and uniformly dispatch schemes for decision-making can be generated in short time.

关 键 词：生态友好型调度 多目标发电优化 差分进化算法 混沌算子 三峡梯级 

分 类 号：TV697[水利工程—水利水电工程]